QIAseq targeted DNA Panel Question

HelenaSC · 11-25-2016, 05:23 AM

Hi alla!
Anyone has been working with QIAseq targeted DNA Panel??

It is our first trial and I'm not very glad with the libraries profiles on Bioanalyzer... We have performed a first experiment with two fresh frozen samples (tubes 1 and 2, input 50 ng DNA) and their FFPE partners (at 2 different Input amount; 100 ng and 250 ng, tubes 3, 4, 5 i 6).

What do you think of the library profile? It seems that there are some large fragments in our libraries...any idea?For the fresh frozen it could be an issue related to the PCR...but in the FFPE?? (It doesn't seem to be beads, in my opinion).

Here are the profiles. Thanks a lot in advance!

jvdmeule · 01-17-2017, 08:17 AM

Hi HelenaSC,

We have just performed a first trial using the QIAseq targeted DNA panel (actionable tumor panel). I also see that we have large fragments in our libraries.
Have you find why these large fragments were present? And have you found a solution?

Many thanks!
Joni

HelenaSC · 01-18-2017, 05:37 AM

Hi Joni,

Yes, the larger fragments are supposed to be overamplification. I made a mistake using the recommendation from the tables, i choosed the wrong conditions regarding the number of cycles in relation to tje number of primers from our panel... Anyway, I repeated the protocol with the correct conditions and, altough some of the samples resulted in much better libraries profile, others just remained the same...

We tried to Run these samples together on a Run and we are experiencing some problems, it seems that the MiSeq was able to detect the clusters (1400) but it wasn't able to give us any more results, and none of the clusters has good passing filter QC...strange... Here is an image from the BySpace, maybe it is more easy to understand...Did you performed a Run?

Imagen1.jpg

Imagen2.jpg

Noemie · 01-25-2017, 06:43 AM

Hi Helena,

We also try to run on Miseq with 8 fresh frozen samples ( input 80 ng DNA) and the MiSeq was able to detect the clusters (1086) and none of the clusters has good passing filter QC!!
Have you found a solution?

Many thanks!

HelenaSC · 01-25-2017, 06:53 AM

Hi!

No news yet...still waiting for QIAgen ansewer to give us some advice...
Did you ckeck the library profile and quantify you samples?

Thanks!

Noemie · 01-25-2017, 07:40 AM

We check the profile and quantify but on bioanalyser DNA 1000 and it was ok.
If we look Data by cycle it's seem to be overclusterised just like you with the same intensity

anineo · 03-20-2017, 12:12 AM

Hi,
have you gotten any answers from Qiagen? Or solved the problem?
We have constructed libraries from FFPE DNA and experience the same problems, lots of fragments around 1000 bp.

(The libraries were diluted 1:5, and nothing was loaded in the last one)

FDKseq · 03-20-2017, 06:17 PM

Hi HelenaSC,

The "larger" fragments in your library was over amplified. The intended library was the peak at the left around 350bp. The over amplified libraries should be the same size as the intended library. They appear as "larger" due to secondary structure since they are not perfectly annealed double strand or just single stranded. Reduce Universal PCR cycle numbers 2-4 should improve over amplification. You will see the increase on intended library peak while decrease on over amplification.

However, over amplification of QIAseq DNA panel won't affect sequencing result, we are doing sequencing on these kind of libraries all time. But, library quantification needs to be accurate and qPCR is preferred when there is over amplification. Sometimes we see library concentration is 2-4 times higher with qPCR quantification than bioanalzyer when there is over amplification.

pirega · 05-30-2018, 07:51 AM

Hi,
We just started to use this kit, at first we had overamplification and some problems during the sequencing in a NextSeq.
Now we are having some issues in the analysis. Is it usual to obtain duplication rates higher than 20%?

anineo · 05-30-2018, 10:49 PM

Hi,
we also had a lot of duplication. We sequenced at 2500x and ended up with a unique depth at around 200x. In our case it may be attributed to low complexity in the sample, as we started with 20 ng cfDNA and possibly even less. However we also eperienced duplication in the tumor DNA samples, which should have been more complex.

How much DNA did you make the libraries from?

pirega · 05-31-2018, 12:08 AM

Hi anineo,
We fixed the problem of overamplification, our library profile is ok. We do not have any experience with cfDNA yet, we are using 125ng of FFPE DNA, is this duplication rate usual?
Thanks

anineo · 05-31-2018, 01:22 AM

Great, how did you fix it? Fewer PCR cycles?

We observed a high duplication rate with FFPE also. Whether it is normal I do not know, but it seems that most studies that use UMIs experience a reduction - which probably just underscores how important it is to include this.

Are you planning to use it on cfDNA, and which kit did you use (which cancer)?

01-17-2017, 08:17 AM	#2
jvdmeule Junior Member Location: Ghent Join Date: Jun 2010 Posts: 1	Hi HelenaSC, We have just performed a first trial using the QIAseq targeted DNA panel (actionable tumor panel). I also see that we have large fragments in our libraries. Have you find why these large fragments were present? And have you found a solution? Many thanks! Joni

01-18-2017, 05:37 AM	#3
HelenaSC Member Location: Spain Join Date: Jun 2013 Posts: 17	Hi Joni, Yes, the larger fragments are supposed to be overamplification. I made a mistake using the recommendation from the tables, i choosed the wrong conditions regarding the number of cycles in relation to tje number of primers from our panel... Anyway, I repeated the protocol with the correct conditions and, altough some of the samples resulted in much better libraries profile, others just remained the same... We tried to Run these samples together on a Run and we are experiencing some problems, it seems that the MiSeq was able to detect the clusters (1400) but it wasn't able to give us any more results, and none of the clusters has good passing filter QC...strange... Here is an image from the BySpace, maybe it is more easy to understand...Did you performed a Run? Imagen1.jpg Imagen2.jpg

01-25-2017, 06:43 AM	#4
Noemie Junior Member Location: Marseille Join Date: Jan 2017 Posts: 2	Hi Helena, We also try to run on Miseq with 8 fresh frozen samples ( input 80 ng DNA) and the MiSeq was able to detect the clusters (1086) and none of the clusters has good passing filter QC!! Have you found a solution? Many thanks!

01-25-2017, 06:53 AM	#5
HelenaSC Member Location: Spain Join Date: Jun 2013 Posts: 17	Hi! No news yet...still waiting for QIAgen ansewer to give us some advice... Did you ckeck the library profile and quantify you samples? Thanks!

01-25-2017, 07:40 AM	#6
Noemie Junior Member Location: Marseille Join Date: Jan 2017 Posts: 2	We check the profile and quantify but on bioanalyser DNA 1000 and it was ok. If we look Data by cycle it's seem to be overclusterised just like you with the same intensity

03-20-2017, 06:17 PM	#8
FDKseq Junior Member Location: maryland Join Date: Mar 2017 Posts: 1	Hi HelenaSC, The "larger" fragments in your library was over amplified. The intended library was the peak at the left around 350bp. The over amplified libraries should be the same size as the intended library. They appear as "larger" due to secondary structure since they are not perfectly annealed double strand or just single stranded. Reduce Universal PCR cycle numbers 2-4 should improve over amplification. You will see the increase on intended library peak while decrease on over amplification. However, over amplification of QIAseq DNA panel won't affect sequencing result, we are doing sequencing on these kind of libraries all time. But, library quantification needs to be accurate and qPCR is preferred when there is over amplification. Sometimes we see library concentration is 2-4 times higher with qPCR quantification than bioanalzyer when there is over amplification.

05-30-2018, 07:51 AM	#9
pirega Junior Member Location: spain Join Date: Apr 2018 Posts: 2	Hi, We just started to use this kit, at first we had overamplification and some problems during the sequencing in a NextSeq. Now we are having some issues in the analysis. Is it usual to obtain duplication rates higher than 20%?

05-30-2018, 10:49 PM	#10
anineo Junior Member Location: Trondheim, Norway Join Date: Mar 2017 Posts: 4	Hi, we also had a lot of duplication. We sequenced at 2500x and ended up with a unique depth at around 200x. In our case it may be attributed to low complexity in the sample, as we started with 20 ng cfDNA and possibly even less. However we also eperienced duplication in the tumor DNA samples, which should have been more complex. How much DNA did you make the libraries from?

05-31-2018, 12:08 AM	#11
pirega Junior Member Location: spain Join Date: Apr 2018 Posts: 2	Hi anineo, We fixed the problem of overamplification, our library profile is ok. We do not have any experience with cfDNA yet, we are using 125ng of FFPE DNA, is this duplication rate usual? Thanks

05-31-2018, 01:22 AM	#12
anineo Junior Member Location: Trondheim, Norway Join Date: Mar 2017 Posts: 4	Great, how did you fix it? Fewer PCR cycles? We observed a high duplication rate with FFPE also. Whether it is normal I do not know, but it seems that most studies that use UMIs experience a reduction - which probably just underscores how important it is to include this. Are you planning to use it on cfDNA, and which kit did you use (which cancer)?